In this project, I train an image classifier to recognize different species of flowers. That is, given an image of a flower, it returns its species. I'm using this dataset of 102 flower categories. A few examples can be seen below.
main.ipynb
contains the project code and documentation. The notebook is designed
to run in any of the following setups by setting the value of a single variable (colab
).
In either case, the dataset is downloaded from AWS and un-zipped.
Local Machine Setup
Fork and clone the repo to a machine which has Anaconda
and PyTorch
and run it.
- Set
colab = False
Google Colabs Setup
Open it from Google Colabs and run it there.
- On Colabs: File > Open notebook... > GITHUB
- In the notebook:
colab = True
- The notebook can be modified in Colabs editor and checked-into Github automatically: File > Save a copy to Github...
- External data: Local Files, Drive, Sheets, and Cloud Storage is an excellent reference for I/O on Google Colabs.